Innovative AI Model Targets Livestock Heart Health
Chase Markel, a Ph.D. student at the University of Wyoming (UW) and a native of Wheatland, is making headlines in the agricultural and scientific communities through his groundbreaking work using artificial intelligence (AI) to address congestive heart failure in cattle. By developing a pioneering AI model trained to predict heart failure risks based on heart images, Markel aims to reduce financial losses incurred by cattle producers due to this condition.
Markel’s innovative tool is the first of its kind, utilizing computer vision technology to assess the risk of heart failure by analyzing key physical indicators in a cow’s heart. His goal is to make early detection more accessible and accurate, ultimately improving animal health and profitability in the beef industry.
Academic Roots and Industry Experience
Having completed both his undergraduate and master’s degrees in UW’s Department of Animal Science, Markel is now pursuing a doctorate in the same department. His research is guided by faculty advisers Hannah Cunningham-Hollinger and Cody Gifford. With a background deeply rooted in the cattle industry, Markel brings firsthand insight to his academic exploration of livestock health.
During his master’s program, he focused on pulmonary hypertension—commonly known as high-altitude disease or brisket disease—in cattle. This condition, which affects animals raised at higher elevations, can lead to congestive heart failure if left untreated. Markel didn’t initially foresee that his research would lead him into the realm of artificial intelligence, but a timely conversation sparked the idea to integrate computing into his work.
From Field to Code: How AI Became a Game-Changer
“I’m not a computer scientist, I’m not an AI guy,” Markel admits. “I’m someone studying heart failure in cattle who happened to make the right connection. It all began with an effort to better understand pulmonary hypertension and heart failure.”
Markel’s research revealed that subclinical cases of pulmonary hypertension—where cattle survive high-altitude disease but still experience health complications—may have a more significant economic impact than previously recognized. These findings fueled his motivation to develop a tool that could detect early-stage heart abnormalities, particularly in the right ventricle. When pressure builds in this part of the heart, it becomes deformed, increasing the risk of heart failure.
By identifying these subtle indicators, Markel believes producers and processing plants can make more informed decisions that benefit the entire supply chain. “Anything we can do to improve traceability and individual animal identification throughout the production cycle is a net positive for the industry,” he says.
Training the Model: Thousands of Images, One Goal
As a fellow at UW’s School of Computing, Markel built an image classification model by training it on nearly 7,000 heart images collected from commercial processing plants in Nebraska and Colorado. The images were scored using a 1-5 system designed by Tim Holt, a collaborator and professor at Colorado State University. Each image was hand-labeled before being used in the machine learning algorithm.
The result? An AI model that can accurately predict the severity of heart failure with 92% accuracy, even when presented with images it has not seen before. While the model is still being refined, it offers a promising proof of concept for using AI in livestock health diagnostics.
Expanding the Scope: Liver Health and Beyond
Encouraged by the success of his heart model, Markel is now developing a similar AI system to evaluate liver images for signs of liver abscesses—another common issue in feedlot cattle. “As researchers, we need to integrate these tools into our research,” he explains. “If we can build the technology, producers can actually use it to improve their bottom line.”
Although the current model is designed for use in processing facilities, Markel envisions future iterations that could directly benefit Wyoming ranchers and other producers in high-altitude regions.
A Vision for the Future of Agricultural Science
Kelly Crane, the Farm Credit Services of America dean in UW’s College of Agriculture, Life Sciences and Natural Resources, applauds Markel’s work. “Chase’s research exemplifies our college’s commitment to Wyoming-relevant research. It bridges emerging technology, producer experience, and faculty expertise to tackle some of the toughest challenges facing our agricultural sector,” Crane states.
Looking ahead, Markel has submitted a provisional patent application to the U.S. Patent and Trademark Office, with hopes of securing full protection by 2026. The patent will safeguard his intellectual property and help ensure the model’s continued development and adoption.
This article is inspired by content from Original Source. It has been rephrased for originality. Images are credited to the original source.